Prediction of the Subgrade Resilient Modulus for the Implementation of the MEPDG in Idaho

This paper focuses on developing two models for subgrade soil characterization for use in the Mechanistic-Empirical Pavement Design Guide (MEPDG). First, a multiple regression model can be used to predict R-value as a function of the soil plasticity index and percent passing No 200 sieve. Second, a Mr predictive model is based on the estimated R-value of the soil. Hence, the models can be used to estimate the Mr value as a Level 2 data input in the MEPDG. For the R-value model development, 8233 records were used from a historical database of Idaho Transportation Department (ITD) laboratory testing on subgrade soils. For the Mr model development, laboratory measured Mr values for different subgrade soils were gathered from literature nationwide. The R-values for these soils were then predicted using the developed R-value model. Hence, a relationship between the Mr and R-value was established. A limited sensitivity analysis was also conducted to address the influence of the material strength upon the predicted distresses by MEPDG.